Operational is essentially querying the data and using that result set as an input to your workflow.
It can also help to view things in two different ways: operational reporting versus analytical reporting. These parts take a few weeks if the data is clean and tiny, months or years if the data is across the enterprise and complex. Where clients and managers have an issue is that all the time saved in querying is actually spent properly modeling the data, writing ETL to conform dimensions and populate facts, and spinning up the reporting solution of SSAS. Even better is that these parameters like $10k, 2014, and 2016 can instantly be changed by clicking a few buttons. These types of queries can instead by manipulated by selecting measures (number of properties pushed to market) and grouping them by dimensions (initial asking price per each $10,000 increment) and also filtering (listed between 20) in a pivot table within Excel. The results eventually get sent back to the client the next day at which point they say, "Cool, but my boss changed his mind and now wants to look at a bigger date range and in a different price range, can you do that?" A SQL report developer has to first take care of their other work, then pound out a complex join on highly normalized tables but only after figuring out the data in each table and dealing with logic issues. Imagine a client wants to see how many properties were listed between date A and B with an initial asking price between Y and Z.
The Mage above is very much correct in associating "data exploration" with OLAP as it lets clients that aren't tech-savvy at all "write" queries in the form of pivot tables in Excel and lets them do it quickly.
0 kommentar(er)
